Introducing uncertainty into pattern discovery in temporal event sequences

Pattern discovery in temporal event sequences is of great importance in many application domains, such as telecommunication network fault analysis. In reality, not every type of event has an accurate timestamp. Some of them, defined as inaccurate events may only have an interval as possible time of occurrence. The existence of inaccurate events may cause uncertainty in event ordering. The traditional support model cannot deal with this uncertainty, which would cause some interesting patterns to be missing. A new concept, precise support, is introduced to evaluate the probability of a pattern contained in a sequence. Based on this new metric, we define the uncertainty model and present an algorithm to discover interesting patterns in the sequence database that has one type of inaccurate event. In our model, the number of types of inaccurate events can be extended to k readily, however, at a cost of increasing computational complexity.

Superstring: A scalable service discovery protocol for the wide-area pervasive environment

Arguably, the world has become one large pervasive computing environment. Our planet is growing a digital skin of a wide array of sensors, hand-held computers, mobile phones, laptops, web services and publicly accessible web-cams. Often, these devices and services are deployed in groups, forming small communities of interacting devices. Service discovery protocols allow processes executing on each device to discover services offered by other devices within the community. These communities can be linked together to form a wide-area pervasive environment, allowing processes in one p u p tu interact with services in another. However, the costs of communication and the protocols by which this communication is mediated in the wide-area differ from those of intra-group, or local-area, communication. Communication is an expensive operation for small, battery powered devices, but it is less expensive for servem and workstations, which have a constant power supply and 81'e connected to high bandwidth networks. This paper introduces Superstring, a peer to-peer service discovery protocol optimised fur use in the wide-area. Its goals are to minimise computation and memory overhead in the face of large numbers of resources. It achieves this memory and computation scalability by distributing the storage cost of service descriptions and the computation cost of queries over multiple resolvers.

Expression-based strategies for discovery of genes involved in testis and ovary development

In recent years, strategies for gene identification based on differential gene expression have become increasingly popular, due in part to the development of microarray technology. These strategies are particularly well suited to the identification of genes involved in sex determination and gonadal development, which unlike the development of other organ systems, proceeds along two very different alternative courses, depending on the sex of the embryo. We have used a high-throughput, array-based expression screen to identify several genes expressed sex-specifically in developing mouse gonads. One of these, vanin 1, appears to play a role in mediating migration of mesonephric cells into the male genital ridge. Progress in characterizing other genes arising from the screen is discussed.

S. cerevisiae as a model for studying mutations in the human gene OPA1 associated with dominant optic atrophy and for drug discovery

L’atrofia ottica dominante (ADOA) è una malattia mitocondriale caratterizzata da difetti visivi, che si manifestano durante l’infanzia, causati da progressiva degenerazione delle cellule gangliari della retina (RGC). ADOA è una malattia genetica associata, nella maggior parte dei casi, a mutazioni nel gene OPA1 che codifica per la GTPasi mitocondriale OPA1, appartenente alla famiglia delle dinamine, principalmente coinvolta nel processo di fusione mitocondriale e nel mantenimento del mtDNA. Finora sono state identificate più di 300 mutazioni patologiche nel gene OPA1. Circa il 50% di queste sono mutazioni missenso, localizzate nel dominio GTPasico, che si pensa agiscano come dominanti negative. Questa classe di mutazioni è associata ad una sindrome più grave nota come “ADOA-plus”. Nel lievito Saccharomyces cerevisiae MGM1 è l’ortologo del gene OPA1: nonostante i due geni abbiano domini funzionali identici le sequenze amminoacidiche sono scarsamente conservate. Questo costituisce una limitazione all’uso del lievito per lo studio e la validazione di mutazioni patologiche nel gene OPA1, infatti solo poche sostituzioni possono essere introdotte e studiate nelle corrispettive posizioni del gene di lievito. Per superare questo ostacolo è stato pertanto costruito un nuovo modello di S. cerevisiae, contenente il gene chimerico MGM1/OPA1, in grado di complementare i difetti OXPHOS del mutante mgm1Δ. Questo gene di fusione contiene una larga parte di sequenza corrispondente al gene OPA1, nella quale è stato inserito un set di nuove mutazioni trovate in pazienti affetti da ADOA e ADOA-plus. La patogenicità di queste mutazioni è stata validata sia caratterizzando i difetti fenotipici associati agli alleli mutati, sia la loro dominanza/recessività nel modello di lievito. A tutt’oggi non è stato identificato alcun trattamento farmacologico per la cura di ADOA e ADOA-plus. Per questa ragione abbiamo utilizzato il nostro modello di lievito per la ricerca di molecole che agiscono come soppressori chimici, ossia composti in grado di ripristinare i difetti fenotipici indotti da mutazioni nel gene OPA1. Attraverso uno screening fenotipico high throughput sono state testate due differenti librerie di composti chimici. Questo approccio, noto con il nome di drug discovery, ha permesso l’identificazione di 23 potenziali molecole attive.

Isolation, characterization and discovery of new alleles of sHsp genes in durum wheat (Triticum durum Desf.)

Gli organismi vegetali mostrano una notevole capacità di adattamento alle condizioni di stress e lo studio delle componenti molecolari alla base dell'adattamento in colture cerealicole di interesse alimentare, come il frumento, è di particolare interesse per lo studio di varietà che consentano una buona produzione con basso input anche in condizioni ambientali non ottimali. L'esposizione delle colture cerealicole a stress termico durante determinate fasi del ciclo vitale influisce negativamente sulla resa e sulla qualità, a questo fine è necessario chiarire le basi genetiche e molecolari della termotolleranza per identificare geni e alleli vantaggiosi da impiegare in programmi di incrocio volti al miglioramento genetico. Numerosi studi dimostrano il coinvolgimento delle sHSP a localizzazione cloroplastica (in frumento sHSP26) nel meccanismo di acquisizione della termotolleranza e la loro interazione con diverse componenti del fotosistema II (PSII) che determinerebbe un’azione protettiva in condizioni di stress termico e altri tipi di stress. Lo scopo del progetto è quello di caratterizzare in frumento duro nuove varianti alleliche correlate alla tolleranza a stress termico mediate l'utilizzo del TILLING (Target Induced Local Lesion In Genome), un approccio di genetica inversa che prevede la mutagenesi e l'identificazione delle mutazioni indotte in siti di interesse. Durante la tesi sono state isolate e caratterizzate 3 sequenze geniche complete per smallHsp26 denominate TdHsp26-A1; TdHsp26-A2; TdHsp26-B1 e un putativo pseudogene denominato TdHsp26-A3. I geni isolati sono stati usati come target in analisi di TILLING in due popolazioni di frumento duro mutagenizzate con EMS (EtilMetanoSulfonato). Nel nostro studio sono stati impiegati due differenti approcci di TILLING: un approccio di TILLING classico mediante screening con High Resolution Melting (HRM) e un approccio innovativo che sfrutta un database di TILLING recentemente sviluppato. La popolazione di mutanti cv. Kronos è stata analizzata per la presenza di mutazioni in tutti e tre i geni individuati mediante ricerca online nel database di TILLING, il quale sfrutta la tecnica dell’exome capture sulla popolazione di TILLING seguito da sequenziamento ad alta processività. Attraverso questa tecnica sono state individuate, nella popolazione mutagenizzata di frumento duro cv. Kronos, 36 linee recanti mutazioni missenso. Contemporaneamente lo screening con HRM, effettuato su 960 genotipi della libreria di TILLING di frumento duro cv. Cham1 ha consentito di individuare mutazioni in una regione di 211bp di interesse funzionale del gene TdHsp26-B1, tra le quali 3 linee mutanti recanti mutazioni missenso in omozigosi. Alcune mutazioni missenso individuate sui due geni TdHsp26-A1 e TdHsp26-B1 sono state confermate in vivo nelle piante delle rispettive linee mutanti generando marcatori codominanti KASP (Kompetitive Allele Specific PCR) con cui è stato possibile verificare anche il grado di zigosità di tali mutazioni. Al fine di ridurre il numero di mutazioni non desiderate nelle linee risultate più interessanti, è stato eseguito il re-incrocio dei mutanti con i relativi parentali wild type ed inoltre sono stati generati alcuni doppi mutanti che consentiranno di comprendere meglio i meccanismi molecolari presieduti da questa classe genica. Gli individui F1 degli incroci sono stati poi genotipizzati con i medesimi marcatori KASP specifici per la mutazione di interesse per verificare la buona riuscita dell’incrocio. Questo approccio ha permesso di individuare ed implementare risorse genetiche utili ad intraprendere studi funzionali relativi al ruolo di smallHSP plastidiche implicate nella acquisizione di termotolleranza in frumento duro e di generare marcatori potenzialmente utili in futuri programmi di breeding.

Data visualization during the early stages of drug discovery

Multidimensional compound optimization is a new paradigm in the drug discovery process, yielding efficiencies during early stages and reducing attrition in the later stages of drug development. The success of this strategy relies heavily on understanding this multidimensional data and extracting useful information from it. This paper demonstrates how principled visualization algorithms can be used to understand and explore a large data set created in the early stages of drug discovery. The experiments presented are performed on a real-world data set comprising biological activity data and some whole-molecular physicochemical properties. Data visualization is a popular way of presenting complex data in a simpler form. We have applied powerful principled visualization methods, such as generative topographic mapping (GTM) and hierarchical GTM (HGTM), to help the domain experts (screening scientists, chemists, biologists, etc.) understand and draw meaningful decisions. We also benchmark these principled methods against relatively better known visualization approaches, principal component analysis (PCA), Sammon's mapping, and self-organizing maps (SOMs), to demonstrate their enhanced power to help the user visualize the large multidimensional data sets one has to deal with during the early stages of the drug discovery process. The results reported clearly show that the GTM and HGTM algorithms allow the user to cluster active compounds for different targets and understand them better than the benchmarks. An interactive software tool supporting these visualization algorithms was provided to the domain experts. The tool facilitates the domain experts by exploration of the projection obtained from the visualization algorithms providing facilities such as parallel coordinate plots, magnification factors, directional curvatures, and integration with industry standard software. © 2006 American Chemical Society.

Toward the discovery of vaccine adjuvants:coupling in silico screening and in vitro analysis of antagonist binding to human and mouse CCR4 receptors

Background Adjuvants enhance or modify an immune response that is made to an antigen. An antagonist of the chemokine CCR4 receptor can display adjuvant-like properties by diminishing the ability of CD4+CD25+ regulatory T cells (Tregs) to down-regulate immune responses. Methodology Here, we have used protein modelling to create a plausible chemokine receptor model with the aim of using virtual screening to identify potential small molecule chemokine antagonists. A combination of homology modelling and molecular docking was used to create a model of the CCR4 receptor in order to investigate potential lead compounds that display antagonistic properties. Three-dimensional structure-based virtual screening of the CCR4 receptor identified 116 small molecules that were calculated to have a high affinity for the receptor; these were tested experimentally for CCR4 antagonism. Fifteen of these small molecules were shown to inhibit specifically CCR4-mediated cell migration, including that of CCR4+ Tregs. Significance Our CCR4 antagonists act as adjuvants augmenting human T cell proliferation in an in vitro immune response model and compound SP50 increases T cell and antibody responses in vivo when combined with vaccine antigens of Mycobacterium tuberculosis and Plasmodium yoelii in mice.

Discovery of a potent phenolic N1-benzylidene- pyridinecarboxamidrazone selective against Gram-positive bacteria

As part of a study into antimycobacterial compounds a set of phenolic N1-benzylidene-pyridinecarboxamidrazones was prepared and evaluated. This report describes the unexpected discovery of a potent compound with a pronounced selectivity for Gram-positive bacteria over Gram-negative micro-organisms. In addition, this compound is active against various drug-resistant Gram-positive bacteria. © 2005 Elsevier Ltd. All rights reserved.

Molecularly imprinted polymers in the drug discovery process

Since molecularly imprinted polymers (MIPs) are designed to have a memory for their molecular templates it is easy to draw parallels with the affinity between biological receptors and their substrates. Could MIPs take the place of natural receptors in the selection of potential drug molecules from synthetic compound libraries? To answer that question this review discusses the results of MIP studies which attempt to emulate natural receptors. In addition the possible use of MIPs to guide a compound library synthesis towards a desired biological activity is highlighted. © 2005 Elsevier B.V. All rights reserved.

Discovery of active proteins directly from combinatorial randomized protein libraries without display, purification or sequencing:identification of novel zinc finger proteins.

We have successfully linked protein library screening directly with the identification of active proteins, without the need for individual purification, display technologies or physical linkage between the protein and its encoding sequence. By using 'MAX' randomization we have rapidly constructed 60 overlapping gene libraries that encode zinc finger proteins, randomized variously at the three principal DNA-contacting residues. Expression and screening of the libraries against five possible target DNA sequences generated data points covering a potential 40,000 individual interactions. Comparative analysis of the resulting data enabled direct identification of active proteins. Accuracy of this library analysis methodology was confirmed by both in vitro and in vivo analyses of identified proteins to yield novel zinc finger proteins that bind to their target sequences with high affinity, as indicated by low nanomolar apparent dissociation constants.

Trepanation:history, discovery, theory

This volume will look at the history of trepanation, the identification of skulls, the tools used to make the cranial openings, and theories as to why trepanation might have been performed many thousands of years ago.

Probabilistic methods for drug discovery

This thesis introduces a flexible visual data exploration framework which combines advanced projection algorithms from the machine learning domain with visual representation techniques developed in the information visualisation domain to help a user to explore and understand effectively large multi-dimensional datasets. The advantage of such a framework to other techniques currently available to the domain experts is that the user is directly involved in the data mining process and advanced machine learning algorithms are employed for better projection. A hierarchical visualisation model guided by a domain expert allows them to obtain an informed segmentation of the input space. Two other components of this thesis exploit properties of these principled probabilistic projection algorithms to develop a guided mixture of local experts algorithm which provides robust prediction and a model to estimate feature saliency simultaneously with the training of a projection algorithm.Local models are useful since a single global model cannot capture the full variability of a heterogeneous data space such as the chemical space. Probabilistic hierarchical visualisation techniques provide an effective soft segmentation of an input space by a visualisation hierarchy whose leaf nodes represent different regions of the input space. We use this soft segmentation to develop a guided mixture of local experts (GME) algorithm which is appropriate for the heterogeneous datasets found in chemoinformatics problems. Moreover, in this approach the domain experts are more involved in the model development process which is suitable for an intuition and domain knowledge driven task such as drug discovery. We also derive a generative topographic mapping (GTM) based data visualisation approach which estimates feature saliency simultaneously with the training of a visualisation model.